Localized Chemistry Department of Chemistry, University of Waterloo

Research

Measurement & Analytical Science

Developing the experiments, instruments, and analysis needed to get structural information from difficult measurements.

200 300 400 500 600 700 Raman shift (cm⁻¹) 100 200 300 400 500 600 700 Temperature (°C) γ-FeOOH γ-Fe₂O₃ α-Fe₂O₃ 0.00 0.25 0.50 0.75 1.00 Normalised intensity 200 300 400 500 600 700 Raman shift (cm⁻¹) 100 200 300 400 500 600 700 Temperature (°C) γ-FeOOH γ-Fe₂O₃ α-Fe₂O₃ 0.00 0.25 0.50 0.75 1.00 Normalised intensity

Variable-temperature Raman spectroscopy of an iron oxyhydroxide: all 303 spectra collected between 20 and 700 °C, shown as a contour map of intensity against Raman shift and temperature. Because the measurement is continuous, it captures the full sequence. Lepidocrocite bands disappear abruptly near 160 °C, the sample passes through a poorly crystalline maghemite phase, and hematite bands appear above ~340 °C. Dashed lines mark the two transitions. Heating under different atmospheres, such as N2, O2, or humidified air, changes the defects present in the final hematite.

Measurement science supports both of our other research areas. Most of the questions we ask are limited by what can be measured, so much of our work goes into developing the experiments, tools, and analysis needed to turn a spectrum into a structural claim.

Disordered materials are especially difficult. Crystallography assumes long-range order, and many functional materials lack it, including amorphous oxides, defective layered materials, and catalysts under operating conditions. The methods below were chosen because they do not depend on long-range order.

Methods we develop and apply

  • X-ray absorption spectroscopy. Element-specific local structure and oxidation state, measured on materials at rest and during operation. We use synchrotron beamlines and our own laboratory X-ray absorption spectrometer.
  • Variable-temperature and in-situ Raman spectroscopy. Analyzing how defects influence thermal response and phase transformations, with Raman mapping to find where in a sample a reaction starts.
  • Electrochemical methods. Spectroelectrochemistry, hydrodynamic methods, Fourier-transformed alternating-current voltammetry, and other techniques that get more information from an electrochemical experiment than a standard current–potential curve.
  • Two-dimensional correlation spectroscopy. Separating overlapping spectral features from phases that coexist during a solid-state reaction, so that each can be assigned and followed with confidence.
  • Structure–property correlation. Statistical analysis across systematically varied samples to link structural descriptors to function, which makes subtle structural effects detectable that a single sample would not show.
  • Machine learning for spectral interpretation. Classification models for large spectral datasets, including customizable models for rapid identification of microplastics by Raman microscopy.

Why it matters

Once a measurement method is published, other groups can apply it to systems we will never work on ourselves. A new method can therefore have a wider impact than the result it was first developed for.